Fair Value Gap Trading Strategy for Futures: Entry, Stop, and When to Pass

Part of our futures prop series. Start with What is a futures prop firm?, then explore: Platforms · Best futures prop firms (2026) · Alpha Futures review (2026) · Performance fee rules explained (2026) · Alpha Group · Futures vs forex prop firms (2026) · Best futures prop firms for beginners (2026) · Fastest payout futures prop firms (2026).

TL;DR: A fair value gap (FVG) is a three-candle pattern. Price jumps, leaves a zone the first and third candles do not overlap, and traders watch that zone if price comes back. On our futures evaluations that box is a location, not a buy button. You still need context, a trigger, a stop that matches structure, and a size that fits our rules: profit target, end-of-day trailing MLL, Daily Loss Guard where it applies, consistency where it applies, and contract caps. If the logical stop does not fit, you pass. We would rather you skip a pretty chart than breach an account.

This is general trading education, not a signal and not personal financial advice. Our evaluations and Qualified Accounts use simulated funds in a simulated trading environment. Simulated results do not reflect real trading outcomes. Performance fees are performance-based. Nothing is guaranteed.

We do not require you to trade FVGs. Plenty of people pass our evaluations with VWAP, opening range, or plain structure. This guide is for traders who already use the pattern and keep blowing the idea up on size, not on the box.

If you only wanted the one-line definition, Alpha Capital has a short glossary: What is a Fair Value Gap (FVG)?. This page is the longer how-to for NQ, ES, MNQ, MES on Alpha Futures.

Start here on our side of the house if you are new to the model: What is a futures prop evaluation? · Futures prop firm rules explained

What fair value gap trading actually is

You mark three candles:

Candle one sets a boundary
Candle two is the jump (people call this displacement)
Candle three does not overlap part of candle one
The leftover interval is the FVG

That is it. You do not need an empty space between two candles sitting next to each other. The middle candle can have traded through those prices. The "gap" is the non-overlap across the three-candle sequence.

Bullish FVG

Strong move up in the middle
Candle one's high sits below candle three's low
You mark from candle one high up to candle three low
Traders may treat a return into that zone as possible support

Bearish FVG

Strong move down in the middle
Candle one's low sits above candle three's high
You mark from candle three high up to candle one low
Traders may treat a return into that zone as possible resistance

Use the wicks. Highs and lows include the wick. If you only draw bodies because it makes a prettier box, you will get different fills than the person who drew wicks, and you will both swear you "traded FVGs."

Words you will see on Twitter and YouTube

We are not asking you to join a cult. You will just bump into these labels, so here is how we use them in this guide:

Displacement: the strong candle that creates the pattern
Mitigation: price has traded back into some or all of the zone
Filled: price has traded through the whole marked zone
Unmitigated: price has not come back yet
Consequent encroachment: the midpoint of the zone (the 50% people limit-order)
Inverse FVG: a failed zone you then watch from the other side. Only if you defined failure first.

None of those are exchange rules. They are trader slang. Write your version in the plan so you do not change the meaning after a red trade.

What an FVG proves, and what it cannot prove

It proves one observable thing: on that timeframe, price moved far enough that candle one and candle three do not overlap.

It does not prove:

that "institutions" left orders in the box
that price must come back to "rebalance"
that the next trade is high probability because a screenshot of last Tuesday worked
that a tight box means tight risk

A candlestick is open, high, low, close. It does not name the buyer. We would rather you treat the FVG as a location and then require a separate reason to click.

On our accounts that second reason has to include risk. A valid FVG that needs a 40-point stop on two NQ is not a valid trade on a $50K evaluation. The chart can still be "right." The account is still dead.

How to mark a bullish FVG

Find a three-candle sequence with a clear push higher in the middle
Compare candle one high with candle three low
Confirm candle one high is below candle three low
Draw from candle one high to candle three low

Example (teaching numbers, not a live level): candle one high 21,210, candle three low 21,214. The bullish FVG is 21,210 to 21,214. Midpoint 21,212.

How to mark a bearish FVG

Mirror image:

Strong push lower in the middle
Candle one low above candle three high
Draw from candle three high to candle one low

Example: candle one low 21,210, candle three high 21,206. Bearish FVG 21,206 to 21,210. Midpoint 21,208.

Borderline gaps (decide this before you journal)

One-minute NQ prints tiny non-overlaps all morning. If a two-tick gap counts when it wins and does not count when it loses, you are not testing a strategy.

Write down, in advance:

minimum zone size in ticks or points
whether the middle candle has to be larger than a recent average
whether you need a close through prior structure, not just a wick
which session counts (we see most NQ/ES day traders care about the New York cash window, not every tick from 6pm ET Sunday)
maximum distance to the next target
whether the next candle is allowed to immediately overlap the zone

There is no official Alpha Futures FVG size. Consistency is the point.

How to build an FVG plan you can actually follow

A box is not a plan. This is the sequence we want on paper before price comes back.

1. Context

One sentence you can prove wrong.

Bad: "The gap should hold."

Better: "I will only look long if 30-minute structure is higher high / higher low, and price returns to the first bullish FVG created after the cash-open break."

Ask yourself:

trending, ranging, or news?
with the higher timeframe or against it?
is this the session I actually trade, or am I bored at 8pm ET?

We do not need you to stack five indicators. We do need a sentence that dies if price does the opposite thing.

2. Qualify the zone before you wait

Not every FVG deserves an alarm.

Useful filters:

middle candle is clearly larger than the last 10 to 20 bars
it printed after a level you marked before the move
it agrees with higher-timeframe direction
it printed in your session
it has not already been fully traded through
there is room to a logical target without needing a miracle

More filters are not automatically smarter. Three filters you can repeat beat eight filters you invent after the loss.

3. Pick one entry model and leave it alone

Model	What you do	What you give up
First touch	Limit at the near edge of the zone	You are in early. No proof the zone is holding.
Midpoint	Limit near 50% of the zone	You wait for more retrace. Price can bounce before you fill.
Confirmation	Lower-timeframe rejection, reclaim, or small structure shift in the zone	More information. Later entry. Often a wider stop.

Pick one for a test block. Switching from confirmation to first touch because "this one looks clean" is how people blow Daily Loss Guard on our Zero plan in twenty minutes.

4. Invalidation (this is not your feelings)

Invalidation is the price action that means the idea is dead.

Common versions:

close through the far side of the FVG
break of the swing that created the directional thesis
no reaction after a set number of bars (time stop)

Your actual stop can sit a small buffer beyond that level for noise and slippage. The buffer is not a licence to "give it more room" after you are already wrong.

5. Size from the stop, not from the account headline

This is the part that matters on our evaluations.

Do not think "$50K account, I can do two NQ." Think: distance to the stop, dollars per contract, remaining room to MLL, and Daily Loss Guard if your plan has one.

Work backwards:

Mark entry
Mark stop
Count ticks or points
Convert to dollars (NQ is not MNQ, ES is not MES)
Choose a quantity that keeps the planned loss inside a personal cap that sits inside our published limits
Include commissions and a bit of slippage in the estimate

If one contract is already too big, you pass, or you use a micro when your plan allows it. Dragging the stop closer so two NQ "fit" is not sizing. That is rewriting the setup.

On our side:

MLL is end-of-day trailing on our current plans. Intraday heat can still feel awful. It does not trail tick by tick the way some other firms do. Read the live article: Futures prop firm rules explained.
Daily Loss Guard applies on some plans (Zero has it on evaluation). Hitting it is a session problem even if MLL is still intact.
Consistency on Advanced and Premium evaluations means one lottery FVG that makes the whole target can still leave you sitting there. Size and frequency both matter.

Confirm live numbers on help.alpha-futures.com before you trade. We update plans. This guide should not be your rule sheet.

6. Target before you are in

Use structure:

prior swing
session high or low
other side of the range
an opposing FVG sitting in front of a level you already cared about

Then look at the ratio. If the stop is honest and the target is 0.4R, maybe the trade is not worth taking. Do not invent a 3R target in empty space because a YouTube video said you need 3R.

A futures example, then the pass we actually want

Teaching numbers. Not a live order. Not a recommendation.

Five-minute Nasdaq, New York morning. Price breaks a range and leaves a bullish FVG from 21,210 to 21,214. Midpoint 21,212.

Written plan:

only with 30-minute structure
enter after a five-minute close back above the midpoint
stop below the swing that started the break, not just below the box
target prior session high
risk no more than the pre-set amount for one idea

Price comes back into the FVG. The swing stop is 14 points away.

Rough dollar math (always check current tick values):

MNQ: $2 per point, so 14 points is about $28 per micro, plus costs
NQ: $20 per point, so 14 points is about $280 per contract, plus costs

On a $50K evaluation, two NQ at $280 planned loss each is a very different day than two MNQ. If the trader's personal cap is $150 for one idea, two NQ is already a no. If Daily Loss Guard is tight, it can be a no even on one NQ.

Pass.

The zone can still bounce. Missing it does not make you "undisciplined." On our evaluations a setup is only valid when the chart idea and the account math both fit. Protecting the attempt is part of execution. See how to pass a futures prop evaluation.

Does this FVG fit our risk limits?

Run this before you send the order:

Daily room: what does the account show if the stop fills, with costs?
MLL room: does this leave enough buffer for a normal losing streak, or is this the last bullet?
Open risk: are you already long MES and about to long MNQ, which is basically one idea?
Consistency: would this size be a freak compared with your last ten trades?
News: is there a release that can gap you through the box?
Re-entry: how many times can this same FVG be traded? Write it. One is a fine answer.

Set a personal daily stop inside whatever hard limit we publish. The goal is to still be thinking at lunch, not staring at a breach email.

Timeframes

There is no "best FVG timeframe." The interval changes what each candle means, so it changes the boxes you see.

Lower timeframes: more FVGs, more decisions, more death by spread and flip-flops
Higher timeframes: fewer, wider zones, often fatter stops

A pairing that a lot of our NQ traders actually use:

30-minute or 15-minute for bias and "does this even matter"
5-minute to mark the FVG and take the trigger
1-minute only if your written plan says so, not because you are scared

Do not scan every interval until one agrees with the trade you already want. That is not multi-timeframe analysis. That is shopping.

Filters (confluence should cut trades, not add size)

With structure. An FVG printed as price breaks a range you already marked is usually cleaner than one in the middle of lunch chop. Define "break" first: wick beyond, close beyond, and hold beyond are three different events.

At a level you already had. Prior day high/low, opening range, VWAP bands if you use them. If you only draw the level after the gap appears, every chart becomes a story.

Volume / order flow. Fine if you write the exact test. "Strong volume" is not a test. "Displacement candle volume above the median of the last 20 bars" is a test.

News. Fast displacement after CPI or FOMC looks gorgeous and fills like garbage. Your plan should say: no new positions inside X minutes, wait a fixed window, or trade news under a separate tested rule set. Do not decide while the candle is still going vertical.

FVG vs similar labels

People mix these up and then argue in Discord.

Label	What it is	How it is not an FVG
Fair value gap	Non-overlap across three candles	Does not need a blank space between two bars
Session gap	Close of one session vs open of the next	Tied to a close and a reopen (think overnight / weekend)
Liquidity void	Fast stretch, often more than three candles	Your definition may span a whole run
Order block	Last opposite candle or base before the impulse	Origin of the move, not the imbalance the move left

Keep your labels stable. If you call a failed FVG an order block after the fact, your journal is fan fiction.

Inverse FVG (only if you pre-define failure)

An inverse FVG starts as a normal FVG that fails under a rule you wrote first. Price closes through it, then later respects it from the other side. Some traders take that flip.

The important word is new. New confirmation, new stop, new size check. Do not rename every loser.

Do fair value gaps "work"?

They can work as a repeatable way to pick a location, an entry, and a stop. That is not the same as "every FVG has edge."

Results move with:

contract (NQ is not CL)
session
timeframe
minimum gap size
trend vs range
first touch vs confirmation
stop and target rules
costs
how you count re-entries

A screenshot proves one pattern existed. It does not tell you the fail rate, the drawdown, or whether costs ate the edge.

The honest question is: does my exact rule set show acceptable results on this market, in this session, after costs, while staying inside our evaluation rules?

How to backtest this without lying to yourself

Write the rules before you hunt charts.

Specify:

instrument and contract month
session and timezone (label ET)
context timeframe and execution timeframe
bullish and bearish formulas
min and max zone size
displacement rule
entry trigger
stop
target and management
news treatment
whether a zone can be traded more than once

If a rule needs "feel," label that judgement and check whether two people would mark the same chart the same way. If they would not, it is not a rule yet.

Record every candidate, including passes:

time it formed
direction and size of the gap
higher-timeframe condition
entry, stop, target
how far it went for you and against you
result after costs
partial fill vs full fill
whether the planned trade even fit the account

Look past win rate. Average win vs average loss, losing streaks, skipped fills, time in trade, results by session. Then test in simulation before you put it on a paid evaluation.

Common mistakes we see

Trading every FVG

Fast markets print dozens of three-candle non-overlaps. Without context you are trading a shape.

Assuming every FVG fills

Price can ignore it, tap the edge, smash through, or never come back. Do not chase the miss.

Entering on the displacement candle

You are often far from invalidation with nothing left to the target. Wait for the entry your plan named.

Moving the stop

If invalidation printed, widening the stop does not improve the thesis. It just makes the loss bigger.

Oversizing a "small" box

The box can be 4 points. The swing stop can be 18. Calculate the real stop in dollars before you pick quantity.

Retrying the same zone

One stop-out becomes three because "it's still fresh." Define mitigation and re-entry before the first attempt.

Changing timeframe after you are in

Dropping to 30 seconds to find a new reason to stay is not management. Manage on the timeframe named in the plan.

FAQs

Does a fair value gap always fill?

No. Quickly, later, partially, or never in the window you trade. Measure fills. Do not bet the evaluation on them.

What is the best FVG timeframe?

None. Pick a context chart and an execution chart that match the contract, the session, and the stop you can actually afford on our plans.

What FVG indicator should I use?

One whose boundaries you understand, with controls for minimum size and fill behaviour. It should save marking time. It should not pick entries or size for you.

Can I use this on an Alpha Futures evaluation?

Yes, as a discretionary method, if you stay inside published Trading Objectives. We do not grade your boxes. We grade whether you hit target without taking the account through MLL (and Daily Loss Guard / consistency where your plan uses them). Confirm live rules before you pay or trade.

Micros or minis?

If the structural stop only fits MNQ or MES, that is the contract. Forcing NQ because "that's what serious traders use" is how evaluations end before lunch.

A cleaner way to use FVGs on our accounts

Mark the zone. Write the context. Pick the trigger. Define invalidation. Size from the stop. Set the target. Write what makes you skip.

Then do that enough times to learn something real, including the skips.

We will not give you an edge by calling every fast candle institutional. If there is an edge, it is the process plus the fact that you still have an account tomorrow.

Your next step: one contract, one session, one context timeframe, one entry model. Record every qualifying FVG, including the ones you pass on, before you change a single rule.

→ Start an Alpha Futures evaluation · How our evaluations work · Our rules, explained